parataxis > hypotaxis Roadmap
Transcript of parataxis > hypotaxis Roadmap
GeorgeWalkden 18/06/2018
Workshoponsyntaxandreconstruc;on 1
University of York
George Walkden Workshop on Syntax and Reconstruction, York, 18.06.2018
Proto-Indo-European:
without Merge? A language
University of York
The context
Comparative and diachronic work with a generative orientation has unquestionably led to a great many discoveries and insights. However, it’s also part of our duty as researchers to consider prominent claims made by scholars of very different theoretical persuasions – especially if those claims seem poorly evidenced, poorly motivated, or simply false. Today’s talk is the early stage of a project to assess the following claim(s):
Proto-Indo-European: a language without Merge? 2 18.06.2018
parataxis > hypotaxis (where “>” is to be read as “precedes”)
University of York
The Parataxis-Precedes-Hypotaxis Hypothesis (PPHH) has a long history: − The term parataxis in its modern sense was introduced by Thiersch (1826) in the context
of historical Greek (opposed to syntaxis there; hypotaxis only in later works) − Very prevalent in historical linguistics before the advent of structuralism
(e.g. Gildersleeve 1883; Delbrück 1900: 411; Small 1924: 125) − Reiterated in more recent works with a functionalist orientation (e.g. Jucker 1991: 203;
Deutscher 2001: ch. 11; Dąbrowska 2015: 230) But almost never explicitly addressed in the generative literature: − Its influence can be seen in O’Neil (1977) and Kiparsky (1995) − Rejected summarily in Roberts (2007: 174–175)
Proto-Indo-European: a language without Merge? 3 18.06.2018
parataxis > hypotaxis
University of York
Roadmap
Part 1: establish what the content of the PPHH is − Spoiler alert: there are several different versions of the PPHH − Only one in principle threatens standard Minimalist assumptions about the architecture of
grammar − That version is very obviously wrong (as far as we can tell) Part 2: empirically evaluate a particular version of the PPHH − Not one that is inherently problematic for generative linguistics if correct − But one that is interesting nonetheless − Precondition for this kind of research: parsed diachronic corpora of various languages
Proto-Indo-European: a language without Merge? 4 18.06.2018
University of York
Content warning
Ideas don’t arise in a vacuum. Some of the ways in which the PPHH is stated (and motivated) in earlier literature make for uncomfortable reading today. − Mitchell (1985) approvingly quotes Small (1924: 125): “It may be laid down as a general
principle that in the progress of language parataxis precedes hypotaxis.” − Small’s following sentence: “The former is associated with the uncultivated mind; the latter,
with the cultivated mind of civilized peoples.” − Andrew (1940: 87): early Old English was characterized by “simply a lack of grammatical
subordination such as we find in the language of children and some primitive people”. This doesn’t mean that (every version of) the PPHH is wrong, of course. But claims (in science as elsewhere) may persist because of ideology rather than merit.
Proto-Indo-European: a language without Merge? 5 18.06.2018 University of York
Flavours of parataxis > hypotaxis
GeorgeWalkden 18/06/2018
Workshoponsyntaxandreconstruc;on 2
University of York
Versions of the PPHH
Harris & Campbell (1995: 284): “in approaching the question of whether hypotaxis develops out of parataxis we encounter the problem that different linguists have in mind different ideas of parataxis, and that at least some of them are vague” A non-exhaustive list: 1. Early human languages lacked Merge. 2. Early human languages lacked self-similar embedding. 3. Early human languages lacked finite subordinate clauses (specifically). 4. Diachronically, hypotactic structures develop out of paratactic structures. 5. Diachronically, adjoined structures replace “embedded” structures. 6. Diachronically, hypotactic structures become more common.
(possible causal factors: “complex” culture/society; Latin; literacy from orality) There may be entailment relations between these hypotheses. But each needs to be considered separately.
Proto-Indo-European: a language without Merge? 7 18.06.2018 University of York
PPHH 1: Early human languages lacked Merge
(aka “The horse, the wheel, and Merge”) − Merge: the operation hypothesized, in Minimalist syntactic theorizing
since Chomsky (1995), to be fundamental to structure building. “Given any two distinct syntactic objects A, B, Merge(A,B) = {A,B}.” (Collins & Stabler 2016: 5)
PPHH 1 can quickly be put to rest: − In a language without an operation such as category-neutral Merge
(defined such that it is able to apply to its own output): “No sentence … could contain more than two words” (Nevins, Pesetsky & Rodrigues 2009: 366)
− Neither Proto-Indo-European nor any other human language has ever been argued to display this property.
− “The cognitive ability to handle finite complementation must have already been a feature of the human brain in the more distant past” (Deutscher 2001: 184–185)
Proto-Indo-European: a language without Merge? 8 18.06.2018
Merge(A,B) = {A,B}
University of York
PPHH 2: Early human languages lacked self-similar embedding
As an empirical claim, this ought to be taken more seriously than PPHH 1. − “although Merge may IN PRINCIPLE combine any two lexical items or phrases an unbounded
number of times, not every imaginable instance of Merge is acceptable in actual languages” (Nevins, Pesetsky & Rodrigues 2009: 366)
− “It is theoretically possible, though unlikely, that some language might be so impoverished in lexical and other resources that only a finite number of non-deviant sentences could be generated by its GP [generative procedure–GW]. If so, it would be a minor curiosity, with no bearing on UG, acquisition, or other significant issues, contrary to much media confusion.” (Chomsky 2013: 35)
− Cf. also Pullum & Scholz (2010): “recursion does not guarantee infinitude”. No grammatical theory I know of predicts that self-similar embedding is a necessary property of human languages. − In the Triggered Merge formalization of Collins & Stabler (2016: section 7), it’s easy to
construct a lexicon where nothing triggers Merge of a constituent of the same type (or of a constituent that could contain a constituent of the same type).
Proto-Indo-European: a language without Merge? 9 18.06.2018 University of York
Are there languages without self-similar embedding?
It depends who you believe. − Futrell et al. (2016), for Pirahã, develop a regular grammar yielding a finite output.
− But this stipulates that “up to 3 instances” of certain elements are allowed. − If instead unbounded repetition is allowed, then they “analyze Pirahã as an infinite
regular language” (2016: 20). − Kornai (2014): “time and again we come across languages where only a finite presentation
seems to make sense”. − But his list includes Akkadian, for which the case has only been made that it lacks
finite complement clauses (Deutscher 2001), and dubious cases such as Proto-Uralic (Ravila 1960).
− “Any finite corpus or set of examples can be given a description as a finite language in principle” (Futrell et al. 2016: 3) – hence, historically attested languages are not a good testing ground.
− Widmer et al. (2017): no language in their sample of 55 Indo-European languages (present or historically attested) lacks NP-within-NP embedding.
No truly compelling case has yet been uncovered.
Proto-Indo-European: a language without Merge? 10 18.06.2018
University of York
PPHH 3: Early human languages lacked finite subordinate clauses (Givón 1979; Karlsson 2009; O’Neil 1977: 207, tentatively)
Assuming that all clauses are CPs, PPHH 2 entails PPHH 3. Givón (1979: 306):
“certain types of languages—those which have only coordination (‘clause chaining’) but no subordination—are found only in preliterate ‘societies of intimates’”
− Givón’s claim is actually stronger: not restricted to finite clauses; uses word “only”.
− Hard to view the Akkadian or Old Assyrian Empire as a non-literate society of intimates, but Akkadian is one of the languages where the case has been best made for lack of finite subordination (Deutscher 2001)
− This claim needs to be assessed with reference to present-day languages; obviously, we have no historical records of non-literate societies pre-20th-century!
− Coordination is a classic instance of self-similar embedding, of course; PPHH 3 could be true without PPHH 2 being true in general.
Working definition of (clausal) subordination:
a CP is subordinated if it is dominated by another CP.
Proto-Indo-European: a language without Merge? 11 18.06.2018 University of York
PPHH 3: Early human languages lacked finite subordinate clauses (Givón 1979; Karlsson 2009; O’Neil 1977: 207, tentatively)
Are there languages without finite subordinate clauses? − Roberts (2007: 174): “the claim that earlier stages of certain languages may have lacked
subordination altogether violates the uniformitarian hypothesis, the idea that all languages at all times reflect the same basic UG, and so cannot be taken seriously in the approach adopted here.” − This of course depends on what we think UG contains. − “On the other hand, it is quite plausible that a language may lack finite clausal
subordination of the familiar type” (2007: 174).
− Delbrück (1900) claims that Proto-Indo-European lacked finite subordination, on the grounds that finite subordinators are not reconstructable. − This is bad reasoning; cf. negation, basic vocabulary − Languages like Mandarin have subordination but no (overt) subordinator − Harris & Campbell (1995: 284): marker/structure fallacy
Proto-Indo-European: a language without Merge? 12 18.06.2018
GeorgeWalkden 18/06/2018
Workshoponsyntaxandreconstruc;on 3
University of York
PPHH 3: some cautionary notes
For PPHH 3 to be correct, there has to be an asymmetry between early and more recently spoken languages in having/lacking finite subordination. − King & Cookson (1890: 204): “We cannot … suppose that hypotaxis is of recent origin in
language; for as far as we can go back in the history of human speech, we find the degradation of sentences to a completely subordinate position fully established.”
− Gildersleeve (1893: xxv): “we have to be on our guard. Hypotaxis is older than our record, and we cannot argue safely as to prehistoric processes”
− Even Karlsson (2009): “Evidence from many language families indicates that non-finite clausal subordination and initial stages of finite clausal subordination existed already in preliterate languages.”
As a categorical claim, PPHH 3 is certainly false. Does it hold statistically? I’m not aware of any studies addressing the question.
Proto-Indo-European: a language without Merge? 13 18.06.2018 University of York
PPHH 4: Diachronically, hypotactic structures develop out of paratactic structures
This is a very different beast. Classic case: reanalysis fusing two independent clauses. [I think that.] [John is here] > [I think [that John is here]] Variant (PPHH 4a), actually very different: reanalysis of adjoined clause as embedded. (Kiparsky 1995, Roberts & Roussou 2003: 116–121) This kind of reanalysis necessarily involves violation of Whitman’s (2000) “conservation of structure” constraint: c-command relations change. Most famous case has been powerfully challenged (Axel-Tober 2017). See Harris & Campbell (1995: 283–310) for sceptical discussion. PPHH 4 is not the focus of this talk.
Proto-Indo-European: a language without Merge? 14 18.06.2018
University of York
PPHH 5: Diachronically, adjoined structures replace embedded structures
Since adjunction is formally a case of self-similar embedding par excellence, and involves (Pair) Merge, PPHH 5 is orthogonal to PPHH 1 and 2. Influential presentation: Kiparsky (1995) for Indo-European. − Proto-Indo-European has only adjoined S; CP is innovated in the history of the subfamilies
(e.g. Germanic). − Wallenberg (2016) presents a supporting quantitative tendency: relative clause
“extraposition” has been getting rarer for centuries in English, Icelandic, French and Portuguese. (Cf. also O’Neil 1977, Suárez-Gómez 2006)
− Whether Kiparsky’s version of PPHH 5 extends beyond Indo-European hasn’t ever really been addressed, to my knowledge.
Today’s results won’t bear directly on PPHH 5.
Proto-Indo-European: a language without Merge? 15 18.06.2018 University of York
Interlude: adjunction in current theories of syntax
Standard mechanism for adjunction in Minimalism: Pair Merge (Chomsky 2001, 2013, Richards 2009, Nomura 2017). Chomsky (2001: 18): − “For structure building, we have so far assumed only the free symmetrical operation
Merge, yielding syntactic operations that are sets, all binary: call them simple. … But is is an empirical fact that there is also an asymmetric operation of adjunction, which takes two objects P and a and forms the ordered pair <a, P>, a adjoined to p. Set-merge and pair-merge are descendants of substitution and adjunction in earlier theories.”
Pair Merge is prima facie a departure from the Strong Minimalist Thesis, and is motivated by the interface with the C-I system, where it can yield predicate composition. In cartographic and Kaynean approaches (Kayne 1994), adjunction is not different from specifier formation, and hence PPHH 4 is not formulable. “It is fair to say that what adjuncts are and how they function grammatically is not well understood.” (Hornstein & Nunes 2008)
Proto-Indo-European: a language without Merge? 16 18.06.2018
University of York
PPHH 6: Diachronically, hypotactic structures become more common
This is possibly the most frequently-encountered version of PPHH in the literature. Usually interpreted with respect to finite clausal subordination in particular (PPHH 6a). Dąbrowska (2015: 230): − “Further telling evidence can be gleaned from historical data. The earliest written texts in a
language are usually highly paratactic … while later texts typically show more use of subordination. The historical increase in the frequency of subordination is gradual”
Karlsson (2009): − “It is a well-known fact that, mainly due to Latin influences, German and English were
syntactically most complex in the 17th century and Swedish in the 19th century” PPHH 6 is a quantitative claim. It can only be assessed using quantitative data from historical corpora.
Proto-Indo-European: a language without Merge? 17 18.06.2018 University of York
Implications of PPHH 6
PPHH 6 has no bearing on questions of grammatical architecture. But it is interesting nonetheless for a variety of reasons. − If PPHH 6 is correct as far as the corpora are concerned, is it a “real change” in the
sense of differences in knowledge of language between generations? − Could in principle be an artefact of the texts available to us from different periods
(poetry, literacy) − Could in principle be a real, but non-linguistic, change
− If it’s a “real change”, and if the causal argument works (see next slide), it indicates that sociocultural factors have an impact on language change (cf. ethnosyntax, Enfield 2002)
− If it’s not a “real change”, it has important implications for the variationist approach to syntactic change (Kroch 1989, Yang 2002, Pintzuk 2003, etc.): how much change in corpus frequency involves change in the weightings of different grammatical options?
But let’s assess the hypothesis first before speculating further! Proto-Indo-European: a language without Merge? 18 18.06.2018
GeorgeWalkden 18/06/2018
Workshoponsyntaxandreconstruc;on 4
University of York
Causal factors
The causal argumentation for all versions of the PPHH has varying levels of quality. Three broad groups of possible factors: − Latin influence (e.g. Karlsson 2009)
− Only really applicable to the Early Modern European written context − Non-finite clauses leading to an increase in finite subordination? − How likely is this to lead to a “real change” in principle?
− Orality > literacy − Chafe (1982) and Biber (1995) show that finite subordinate clauses are more
common in spoken than in written texts − But very difficult to disentangle “real change” from genre effects; cf. present day
− Cultural complexity & communicative needs (e.g. Givón 1979; Deutscher 2001) − Difficult to find a robust proxy; different notions of “complexity”/“needs” − Important to avoid discredited notions of “primitiveness” and lower intelligence − Causal chain rarely made explicit (though see Deutscher 2001: 166–186)
Proto-Indo-European: a language without Merge? 19 18.06.2018 University of York
Overview of hypotheses
Categorical Tendency Structures No Merge > Merge
PPHH 1 ? ?
No self-embedding > self-embedding
PPHH 2 PPHH 6 ?
No subordination > finite subordination
PPHH 3 PPHH 6A PPHH 4
Clausal adjunction > embedding
PPHH 5 ? PPHH 4A
Proto-Indo-European: a language without Merge? 20 18.06.2018
In the rest of today’s talk: PPHH 6a will be the focus.
University of York
Does finite clausal
more common over time? subordination become
University of York
Methods: investigating PPHH 6a
Crucially relies on availability of parsed diachronic corpora. “Hypotaxis coefficient”: proportion of finite clauses that are subordinate/embedded. − Finite unembedded clauses: IP-MAT* in Penn-style parsed corpora
(includes e.g. imperatives, exclamatives, coordinated clauses) − Finite subordinate/embedded clauses: basically IP-SUB* in Penn-style parsed corpora
(includes e.g. relatives, complement clauses, adverbial clauses) − Some variation in how interrogatives are treated – ask me if interested
(shouldn’t affect the overall results much)
Languages investigated: English, Icelandic, French, Portuguese, Irish, Chinese Non-finite clauses left out of consideration entirely.
Proto-Indo-European: a language without Merge? 22 18.06.2018
University of York
English
Proto-Indo-European: a language without Merge? 23 18.06.2018
0.00
0.25
0.50
0.75
1.00
750 1000 1250 1500 1750Date
Hyp
otax
is c
oeffi
cien
t
Size5000
10000
15000
− YCOE (Taylor et al. 2003)
− PPCME2 (Kroch & Taylor 2000)
− PPCEME (Kroch et al. 2005)
− PPCMBE (Kroch et al. 2010)
University of York
English with genre
Proto-Indo-European: a language without Merge? 24 18.06.2018
− “Non-fiction” (purple) is something of a dustbin category.
− Legal texts high; diaries and bibles low
0.00
0.25
0.50
0.75
1.00
750 1000 1250 1500 1750Date
Hyp
otax
is c
oeffi
cien
t
GenreBible
Diaries
Drama
Fiction
Legal
Letters
Non-fiction
Sermon
Size5000
10000
15000
GeorgeWalkden 18/06/2018
Workshoponsyntaxandreconstruc;on 5
University of York
Icelandic
Proto-Indo-European: a language without Merge? 25 18.06.2018
0.00
0.25
0.50
0.75
1.00
1200 1400 1600 1800 2000Date
Hyp
otax
is c
oeffi
cien
t
GenreBible
Fiction
Legal
Non-fiction
Religious
ReligiousSaga
Saga
Sermon
Size1000
2000
3000
4000
5000
− IcePaHC (Wallenberg et al. 2011)
− Sagas typically have less than average hypotaxis
University of York
French
Proto-Indo-European: a language without Merge? 26 18.06.2018
0.00
0.25
0.50
0.75
1.00
1200 1400 1600Date
Hyp
otax
is c
oeffi
cien
t
GenreDiaries
FictionMixed
FictionProse
FictionVerse
Letters
Non-fiction
Sermon
Size5000
10000
15000
20000
− MCVF (Martineau et al. 2010)
− Apparent early rise is exclusively due to dominance of verse texts in this period
University of York
Portuguese
Proto-Indo-European: a language without Merge? 27 18.06.2018
0.00
0.25
0.50
0.75
1.00
1500 1600 1700 1800Date
Hyp
otax
is c
oeffi
cien
t
GenreDissertation
Drama
Grammar
Letters
Narrative
News
Records
Size2000
4000
6000
8000
− Tycho Brahe Corpus (Galves, Andrade & Faria 2017)
− News texts & dramas typically low
University of York
Old and Middle Irish
Proto-Indo-European: a language without Merge? 28 18.06.2018
0.00
0.25
0.50
0.75
1.00
700 800 900 1000 1100Date
Hyp
otax
is c
oeffi
cien
t
Size300
600
900
1200
GenreLegal
Narrative
Religious
Sermon
− Parsed Corpus of Old and Middle Irish (Lash 2014)
− Hard to generalize about genre
University of York
Chinese
Proto-Indo-European: a language without Merge? 29 18.06.2018
− ChiParHC (Li 2017) − Again, hard to
generalize about genre
0.00
0.25
0.50
0.75
1.00
0 500 1000 1500Date
Hyp
otax
is c
oeffi
cien
t
Size500
1000
1500
2000
GenreHandbook
Legal
Novel
Philosophy
Politics
University of York
Evaluation
Mixed-effects binomial logistic regression using R and lme4 package − Fixed effect: date − Random intercept: text Linear effect of time should at least be detectable if the hypothesis is correct! Nagelkerke R2, a measure of goodness of fit, calculated using Nakagawa & Schielzeth (2013) method and MuMIn R package. Gives percentage of variance explained by the model. − Marginal R2: only fixed effects (date) − Conditional R2: fixed and random effects (date and text)
Proto-Indo-European: a language without Merge? 30 18.06.2018
English Icelandic French Portuguese Irish Chinese Log odds -0.00022 0.00014 -0.00141 0.00326 0.00092 -0.00016
Marginal R2 0.1% 0.0% 1.5% 2.5% 0.4% 0.3% Conditional R2 12.4% 6.7% 9.9% 4.7% 2.6% 3.0%
GeorgeWalkden 18/06/2018
Workshoponsyntaxandreconstruc;on 6
University of York
Overview
Proto-Indo-European: a language without Merge? 31 18.06.2018
0.00
0.25
0.50
0.75
1.00
750 1000 1250 1500 1750Date
Hyp
otax
is c
oeffi
cien
t
Size5000
10000
15000
0.00
0.25
0.50
0.75
1.00
1200 1400 1600 1800 2000Date
Hyp
otax
is c
oeffi
cien
t Size1000
2000
3000
4000
5000
0.00
0.25
0.50
0.75
1.00
1200 1400 1600Date
Hyp
otax
is c
oeffi
cien
t
Size5000
10000
15000
20000
0.00
0.25
0.50
0.75
1.00
1500 1600 1700 1800Date
Hyp
otax
is c
oeffi
cien
t
Size2000
4000
6000
8000
English Icelandic French Portuguese Irish Chinese
No robust support for any version of the PPHH. − English, Icelandic, Irish, Chinese: no consistent direction of change. − French: increase in hypotaxis 1100–1200, but early texts are in verse. − Portuguese: gentle but steady decrease in hypotaxis over the timespan of the corpus. Does genre play a role? Yes, but irrelevant to the hypothesis as far as we can tell. − The most hypotactic texts in English are legal texts. − A consistent role for genre is exactly what we’d predict given Chafe’s (1982) and Biber’s
(1995) results, if performance effects are constant. − So unless the corpora are unbalanced and genre effects are counteracting a real
diachronic trend, the result basically stands.
0.00
0.25
0.50
0.75
1.00
700 800 900 1000 1100Date
Hyp
otax
is c
oeffi
cien
t
Size300
600
900
1200
0.00
0.25
0.50
0.75
1.00
0 500 1000 1500Date
Hyp
otax
is c
oeffi
cien
t
Size500
1000
1500
2000
University of York
Bonus languages: Latin, Slavic/Russian, Georgian
These corpora don’t have constituency parsing. − Latin: PROIEL − Slavic/Russian: PROIEL − Georgian: Georgian National Corpus Approximation to the hypotaxis coefficient: number of overt subordinators divided by the number of finite verbs. This seems to work reasonably well (more testing needed). Proto-Indo-European: a language without Merge? 32 18.06.2018
University of York
Latin
Proto-Indo-European: a language without Merge? 33 18.06.2018
− PROIEL (Haug & Jøhndal 2008)
− Again, hard to generalize about genre
0.00
0.25
0.50
0.75
1.00
-200 0 200 400Date
Hyp
otax
is c
oeffi
cien
t
GenreBible
Drama
Fiction
Letters
Non-fiction
Size5000
10000
15000
20000
University of York
Slavic/Russian
Proto-Indo-European: a language without Merge? 34 18.06.2018
− PROIEL (Haug & Jøhndal 2008)
− Bible texts are Old Church Slavonic; narrative texts are Russian
− Too little here to say anything meaningful at all
0.00
0.25
0.50
0.75
1.00
1000 1200 1400Date
Hyp
otax
is c
oeffi
cien
t
GenreBible
Narrative
Size2500
5000
7500
10000
University of York
Georgian
Proto-Indo-European: a language without Merge? 35 18.06.2018
− Georgian National Corpus (Gippert & Tandashvili 2015)
− Philosophical and legal texts most hypotactic
0.00
0.25
0.50
0.75
1.00
800 1200 1600Date
Hyp
otax
is c
oeffi
cien
t
Size10000
20000
30000
40000
50000
GenreAnachoretic
Bible
Commentary
Hagiography
History
Legal
Philosophy
Religion
Science
Sermon
SpiritualVerse
University of York
Conclusion
• It’s widely agreed that parataxis > hypotaxis. Much less widely agreed what this actually means.
• Focusing on the idea that finite clausal subordination becomes more prevalent over time, I have found no support for this in parsed diachronic corpora of English, Icelandic, French, Portuguese, Irish, or Chinese. − Maybe the corpus annotation is wrong. − Maybe the choice of languages is wrong. − But insofar as parataxis > hypotaxis is an empirical question,
the burden of proof should be shifting at least somewhat. • Much future work suggests itself:
− More languages. More consideration of genre. What about non-finite clauses? − Suggestions welcomed!
Proto-Indo-European: a language without Merge? 36 18.06.2018
Thank you for your attention!
GeorgeWalkden 18/06/2018
Workshoponsyntaxandreconstruc;on 7
University of York
References (1)
− Andrew, S. O. 1940. Syntax and style in
Old English. Cambridge: Cambridge University Press.
− Axel-Tober, Katrin. 2017. The development of the declarative complementizer in German. Language: Historical Syntax 93, e29–e65.
− Biber, Douglas. 1995. Dimensions of register variation: a cross-linguistic comparison. Cambridge: Cambridge University Press.
− Chafe, Wallace. 1982. Integration and involvement in speaking, writing, and oral literature. In Deborah Tannen (ed.), Spoken and written language: Exploring orality and literacy, 35–53. Norwood, NJ: Ablex.
− Chomsky, Noam. 1995. The Minimalist Program. Cambridge, MA: MIT Press.
− Chomsky, Noam. 2001. Beyond
explanatory adequacy. MIT Working Papers in Linguistics 20, 1–28.
− Chomsky, Noam. 2013. Problems of projection. Lingua 130, 33–49.
− Collins, Chris, & Edward Stabler. 2016. A formalization of Minimalist syntax. Syntax 19, 43–78.
− Dąbrowska, Ewa. 2015. Language in the mind and in the community. In Jocelyne Daems, Eline Zenner, Kris Heylen, Dirk Speelman & Hubert Cuyckens (eds.), Change of paradigms – new paradoxes: recontextualizing language and linguistics, 221–236. Berlin: de Gruyter.
− Delbrück, Berthold. 1900. Vergleichende Syntax der indogermanischen Sprachen. Vol. 3. Strasbourg: Karl J. Trübner.
Proto-Indo-European: a language without Merge? 37 18.06.2018 University of York
References (2)
− Deutscher, Guy. 2001. Syntactic change
in Akkadian: the evolution of sentential complementation. Oxford: Oxford University Press.
− Enfield, Nicholas J. (ed.) 2002. Ethnosyntax: explorations in grammar and culture. Oxford: Oxford University Press.
− Futrell, Richard, Laura Stearns, Daniel L. Everett, Steven T. Piantadosi & Edward Gibson. 2016. A corpus investigation of syntactic embedding in Pirahã. PLoS ONE 11, 1–20.
− Galves, Charlotte, Aroldo L. de Andrade, & Pablo Faria. 2017. Tycho Brahe Parsed Corpus of Historical Portuguese.
− Gildersleeve, Basil L. 1883. On the final sentence in Greek. The American Journal of Philology 4, 416–444.
− Gildersleeve, Basil L. 1893. Some
problems in Greek syntax. Transactions of the American Philological Association 24, xxiv–xxvii.
− Gippert, Jost, & Manana Tandashvili. 2015. Structuring a diachronic corpus: the Georgian National Corpus project. In Jost, Gippert & Ralf Gehrke (eds.), Historical corpora: challenges and perspectives, 305–322. Tübingen: Narr.
− Givón, Talmy. 1979. From discourse to syntax: Grammar as a processing strategy. In Talmy Givón (ed.), Syntax and semantics, vol. 12: Discourse and syntax, 81–112. New York, NY: Academic Press.
− Harris, Alice C., & Lyle Campbell. 1995. Historical syntax in cross-linguistic perspective. Cambridge: Cambridge University Press.
Proto-Indo-European: a language without Merge? 38 18.06.2018
University of York
References (3)
− Haug, Dag T. T., & Marius L. Jøhndal.
2008. Creating a parallel treebank of the old Indo-European bible translations. In Caroline Sporleder & Kiril Ribarov (eds.). Proceedings of LaTeCH 2008, 27–34.
− Hornstein, Norbert, & Jairo Nunes. 2008. Adjunction, labeling, and Bare Phrase Structure. Biolinguistics 2, 57–86.
− Jucker, Andreas H. 1991. Between hypotaxis and parataxis: clauses of reason in Ancrene Wisse. In Dieter Kastovsky (ed.), Historical English syntax, 203–219. Berlin: de Gruyter.
− Karlsson, Fred. 2009. Origin and maintenance of clausal embedding complexity. In Geoffrey Sampson, David Gil & Peter Trudgill (eds.), Language complexity as an evolving variable, 192–202. Oxford: Oxford University Press.
− Kayne, Richard S. 1994. The
antisymmetry of syntax. Cambridge, MA: MIT Press.
− King, John E., & Christopher Cookson. 1890. An introduction to the comparative grammar of Greek and Latin. Oxford: Clarendon.
− Kiparsky, Paul. 1995. Indo-European origins of Germanic syntax. In Adrian Battye & Ian Roberts (eds.), Clause structure and language change, 140–169. Oxford: Oxford University Press.
− Kornai, András. 2014. Resolving the infinitude controversy. Journal of Logic, Language and Information 23, 481–492.
− Kroch, Anthony. 1989. Reflexes of grammar in patterns of language change. Language Variation & Change 1, 199–244.
Proto-Indo-European: a language without Merge? 39 18.06.2018 University of York
References (4)
− Kroch, Anthony, & Ann Taylor. 2000. The
Penn-Helsinki Parsed Corpus of Middle English (PPCME2).
− Kroch, Anthony, Beatrice Santorini, & Lauren Delfs. 2004. The Penn-Helsinki Parsed Corpus of Early Modern English (PPCEME).
− Kroch, Anthony, Beatrice Santorini, & Ariel Diertani. 2010. The Penn Parsed Corpus of Modern British English (PPCMBE).
− Lash, Elliott. 2014. The Parsed Old and Middle Irish Corpus (POMIC). Version 0.1.
− Li, Man. 2017. Chinese Parsed Historical Corpus (ChiParHC).
− Martineau, France, Paul Hirschbühler, Anthony Kroch, & Yves Charles Morin. 2010. Modéliser le changement: Les voies du français.
− Mitchell, Bruce. 1985. Old English syntax.
2 vols. Oxford: Clarendon. − Nakagawa, Shinichi, & Holger Schielzeth.
2013. A general and simple method for obtaining R2 from generalized linear mixed-effects models. Methods in Ecology and Evolution 4, 133–142.
− Nevins, Andrew, David Pesetsky, & Cilene Rodrigues. 2009. Pirahã exceptionality: a reassessment. Language 85, 355–404.
− Nomura, Masashi. 2017. Pair-merge and feature-valuation via minimal search: evidence from Icelandic. In Aaron Kaplan, Abby Kaplan, Miranda K. McCarvel, & Edward J. Rubin (eds.), Proceedings of the 34th West Coast Conference on Formal Linguistics, 395–403. Somerville, MA: Cascadilla Proceedings Project.
Proto-Indo-European: a language without Merge? 40 18.06.2018
University of York
References (5)
− O’Neil, Wayne. 1977. Clause adjunction in
Old English. General Linguistics 17, 199–211.
− Pintzuk, Susan. 2003. Variationist approaches to syntactic change. In Brian Joseph & Richard Janda (eds.), The handbook of historical linguistics, 509–528. Oxford: Blackwell.
− Pullum, Geoffrey K., & Barbara Scholz. 2010. Recursion and the infinitude claim. In Harry van der Hulst (ed.), Recursion and human language, 113–137. Berlin: de Gruyter.
− Ravila, Paavo. 1960. Proto-Uralic. In B. Collinder (ed.), Comparative grammar of the Uralic languages, 250–251. Stockholm: Almqvist & Wiksell.
− Richards, Marc. 2009. Internal pair-
merge: the missing mode of movement. Catalan Journal of Linguistics 8, 55–73.
− Roberts, Ian. 2007. Diachronic syntax. Oxford: Oxford University Press.
− Small, George W. 1924. The comparison of inequality. Baltimore: University Press.
− Suárez-Gómez, Cristina. 2006. Relativization in early English (950–1250): the position of relative clauses. Berlin: Peter Lang.
− Taylor, Ann, Anthony Warner; Susan Pintzuk, & Frank Beths. 2003. The YorkToronto-Helsinki Parsed Corpus of Old English Prose.
− Thiersch, Friedrich. 1826. Griechische Grammatik, vorzüglich des homerischen Dialekts. 3rd edition. Leipzig: Fleischer.
Proto-Indo-European: a language without Merge? 41 18.06.2018 University of York
References (6)
− Wallenberg, Joel C. 2016. Extraposition is
disappearing. Language: Historical Syntax 92, e237–e256.
− Wallenberg, Joel C., Anton K. Ingason, Einar F. Sigurðsson, & Eiríkur Rögnvaldsson. 2011. Icelandic Parsed Historical Corpus (IcePaHC). Version 0.9.
− Whitman, John. 2000. Relabelling. In Susan Pintzuk, George Tsoulas & Anthony Warner (eds.), Diachronic syntax: models and mechanisms, 220–238. Oxford: Oxford University Press.
− Widmer, Manuel, Sandra Auderset, Johanna Nichols, Paul Widmer & Balthasar Bickel. 2017. NP recursion over time: evidence from Indo-European. Language 93, 799–826.
Proto-Indo-European: a language without Merge? 42 18.06.2018
− Yang, Charles D. 2002. Knowledge and
learning in natural language. Oxford: Oxford University Press.